The plateau conversation almost always starts with a number that is not a plateau. Three weeks, sometimes two. The scale has not moved, and the conclusion has already been reached that the drug has stopped working.
Weight is a noisy measurement of a slow process. Distinguishing signal from noise is most of the skill here.
What the trial curves actually look like
Worth knowing, because they set the expectation.
In STEP 1, semaglutide 2.4 mg produced roughly 15 percent average weight loss over 68 weeks, and the curve had not fully flattened when the trial ended. In SURMOUNT-1, tirzepatide at 15 mg produced around 21 percent over 72 weeks, on a similar trajectory.
Two implications. The process is long — the bulk of the loss happens after month four, not before it. And the flattening arrives late, typically past the one-year mark. A stall at month five is very unlikely to be your final answer.
Week 60+
Point at which the average STEP 1 curve began flattening
NEJM, 2021
8–12 weeks
Duration of no trend change required before calling a genuine plateau
~5 lb
Typical range the scale moves from water and glycogen alone, independent of fat
Why a real plateau happens
Your body is smaller. This is the largest and least interesting reason. A person eighty pounds lighter burns meaningfully fewer calories at rest and in movement. The intake that created a deficit at 260 pounds is closer to maintenance at 180.
Metabolic adaptation. Beyond what size alone predicts, resting expenditure falls a little further during sustained weight loss — the body defending its previous weight. The magnitude is debated and the direction is not.
Intake has drifted. Not through any failure of will. Appetite suppression is strongest early; the gut adapts; food becomes appealing again in month eight in a way it was not in month two. Almost nobody notices this happening, and almost everybody underestimates intake — the effect is well documented and largely unconscious.
Muscle has been lost. If a quarter to a third of what you have lost was lean tissue, you have reduced your own metabolic floor. This is the one that compounds, and the one the muscle question exists to address.
The dose is behind you. If you stopped escalating at 1.0 mg because week five was unpleasant, you are being treated below the range where most of the trial effect occurred.
The four things to check first
1. Are you actually at a therapeutic dose? Many people stall at an intermediate step and assume the drug has failed. If you are at 1.0 mg of semaglutide or 7.5 mg of tirzepatide and tolerating it well, there is ladder left — see the dose charts.
2. Is your protein where it should be? Around 1.2 to 1.6 g per kilogram. Under-eating protein while losing weight rapidly is the fastest route to losing muscle and lowering your own expenditure. The practical version is in protein on a GLP-1 and what to eat.
3. Are you lifting anything? Resistance training two or three times a week is the only intervention that meaningfully protects lean mass. Creatine is the one supplement with real evidence behind it here — see the best creatine for GLP-1 users and supplements for lifting on a GLP-1. It also frequently produces a scale that stalls while the mirror improves — which is a good outcome misread as a bad one. See the resistance training minimum.
4. Has intake genuinely drifted? Track everything for one honest week. Not to diet — to find out. Most people are surprised, and the surprise is usually liquid calories, evening grazing, or portions that crept back as the food noise returned.
A scale that stops moving while your waist keeps shrinking is not a plateau. It is body composition, and it is the outcome you actually wanted.
The measurements that catch what the scale misses
Waist circumference, monthly, same place. Clothes. Photographs in the same light. Strength in the gym.
Someone gaining two pounds of muscle while losing two pounds of fat reads as a total failure on a bathroom scale and as excellent progress on every other measure. This is common in people who start lifting in month four, which is exactly when the first stall tends to arrive.
When a plateau is the destination
Here is the reframe that matters most, and it is the least welcome.
Weight loss is supposed to stop. The goal was never an indefinite descent. If you have lost 18 percent of your body weight and the curve has flattened at a weight you can sustain, you have not failed — you have finished the losing phase and started the harder one.
The conversation to have at that point is about maintenance, which almost nobody schedules in advance, and about whether and how you would ever come off.
What not to do
Do not skip doses to reset anything. There is nothing to reset, and the sawtooth costs you weeks — see missed doses.
Do not cut calories dramatically. Deepening the deficit on top of appetite suppression accelerates muscle loss and metabolic adaptation, which makes the plateau worse rather than better.
Do not switch drugs immediately. Switching means restarting at the bottom of a new ladder and losing four to five months of dose position — worth it for a genuine non-response, wasteful for a six-week stall. The mechanics are in switching between GLP-1s.
The honest arithmetic
Most plateaus are eight weeks of ordinary variation being read as a verdict.
Of the ones that are real, most resolve with a dose that finished climbing, protein that hit its target, and two sessions a week under a barbell. Very few are the drug having stopped working — and you cannot tell which you have without a record that goes back three months.